Classification-Based Approach to Concept Map Generation in Adaptive Learning

Data mining has recently drawn a lot of interests as an effective way of generating a concept map in an adaptive learning system that provides students with the personalized learning guidance. Even with significant progresses witnessed in this field, the data mining-based concept map generation needs further improvement both in accuracy and complexity before it can be employed in actual education services. This paper proposes a classification-based approach to significantly reduce computational complexity of concept map generation while maintaining the accuracy of the generated concept map, and demonstrates through simulations that the approach accomplishes the objectives.